U.S. patent application number 11/945138 was filed with the patent office on 2008-04-24 for method of providing product database.
This patent application is currently assigned to NHN CORPORATION. Invention is credited to HYANG CHEOL LEE, JOON HYOCK MOON, BUM JOO PARK.
Application Number | 20080097976 11/945138 |
Document ID | / |
Family ID | 37306544 |
Filed Date | 2008-04-24 |
United States Patent
Application |
20080097976 |
Kind Code |
A1 |
LEE; HYANG CHEOL ; et
al. |
April 24, 2008 |
METHOD OF PROVIDING PRODUCT DATABASE
Abstract
A category recommendation method includes maintaining a category
database including an internal identifier and an internal category
to which the internal identifier belongs; externally receiving a
foreign identifier and a foreign category where the foreign
identifier belongs to from an outside; extracting a query
associated with the foreign identifier; and recommending at least
one of the internal categories stored in the category database by
using the query as a category corresponding to the foreign
identifier.
Inventors: |
LEE; HYANG CHEOL; (SEOUL,
KR) ; MOON; JOON HYOCK; (SEOUL, KR) ; PARK;
BUM JOO; (SEOUL, KR) |
Correspondence
Address: |
KNOBBE MARTENS OLSON & BEAR LLP
2040 MAIN STREET
FOURTEENTH FLOOR
IRVINE
CA
92614
US
|
Assignee: |
NHN CORPORATION
Seongnam-si
KR
|
Family ID: |
37306544 |
Appl. No.: |
11/945138 |
Filed: |
November 26, 2007 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
PCT/KR2006/001858 |
May 18, 2006 |
|
|
|
11945138 |
Nov 26, 2007 |
|
|
|
Current U.S.
Class: |
1/1 ;
707/999.004; 707/E17.062; 707/E17.066; 707/E17.109 |
Current CPC
Class: |
G06Q 30/0633 20130101;
G06F 16/3322 20190101; G06F 16/9535 20190101 |
Class at
Publication: |
707/004 ;
707/E17.062 |
International
Class: |
G06F 17/30 20060101
G06F017/30 |
Foreign Application Data
Date |
Code |
Application Number |
May 26, 2005 |
KR |
10-2005-0044404 |
Claims
1. A method of updating a product database, comprising: providing a
database comprising a plurality of database entries, wherein each
database entry comprises a classification field and a plurality of
additional fields, wherein the classification field comprises a
classification of an item of the database entry; providing data
associated with a product item for adding to the database as a new
entry; generating a query from the data associated with the
product; computing a correlation value of the query with
information saved in at least one selected field of a database
entry, wherein computing is repeated for at least part of the
plurality of database entries; identifying database entries
including information in the at least one selected field, which has
a correlation value greater than a predetermined value; determining
the most popular classification among classifications saved in the
classification fields of the identified database entries; and
assigning the most popular classification to the classification
field of the new entry.
2. The method of claim 1, wherein the data is retrieved from an
on-line store for selling the product.
3. The method of claim 2, wherein the query comprises a word or
phrase that the on-line store uses to identify, describe or
classify the product.
4. The method of claim 2, wherein the plurality of fields further
comprises a product source field, wherein the method further
comprises adding to the product source field a name of the on-line
store.
5. The method of claim 2, wherein the plurality of fields further
comprises a price field, wherein the method further comprises
adding to the price field a price of the product available at the
on-line store.
6. The method of claim 5, further comprising filtering the price
field of a group of database entries using a price range so as to
select a smaller group of entries with a price falling in the price
range.
7. The method of claim 1, providing data comprises receiving the
data from an on-line store, wherein the data comprises at least one
selected from the group consisting of a name of the product, a
classification of the product, an identification code of the
product, and a description of the product that are used in the
on-line store.
8. The method of claim 7, wherein the query is created using terms
in the name of the product, the classification of the product, the
identification code of the product, and the description of the
product.
9. The method of claim 7, wherein the plurality of fields comprise
a product identification code field, wherein the method further
comprises determining whether the database comprises a database
entry having an internal identifier in the product identification
code field that is identical to the identification code of the
product that is used in the on-line store.
10. The method of claim 9, wherein generating, computing,
identifying and determining are conducted only when it is
determined that there is no database entry that includes the
identical identification code in the product identification code
field.
11. The method of claim 9, wherein the at least one selected field
comprises the product identification code field.
12. The method of claim 9, wherein the at least one selected field
comprises a field other than the product identification code
field.
13. The method of claim 9, wherein the internal identifier is a
model number of the product.
14. The method of claim 1, wherein the query comprises a string of
plurality of words associated with the product.
15. The method of claim 1, wherein determining comprise:
identifying different classifications used in the identified
database entries; counting the number of database entries for each
different classification; and comparing the counted numbers to
determine the most popular classification.
16. A method of running a price comparison engine, the method
comprising: maintaining a product database comprising the method of
claim 1; receiving, from a remote computer, a search request for a
type of products; conducting a search in the database using
information included in the search request; and sending a search
result to the remote computer.
17. A method of a product database, comprising: providing a product
database comprising a plurality of fields for each database entry,
the plurality of fields comprising a classification field;
providing data associated with a product for adding to the database
as a new entry; extracting at least one term from the data
associated with the product; conducting at least one search using
the at least one term in the database so as to locate a plurality
of entries including the at least one term in at least one field
thereof; and assigning, to the classification field of the new
entry, information or a value that is the most popular in the
classification fields of the plurality of entries located in the at
least one search.
18. The method of claim 17, wherein the data is retrieved from an
on-line store for selling the product.
19. The method of claim 18, wherein the data comprises at least one
selected from the group consisting of a name of the product, a
classification of the product, an identification code of the
product, and a description of the product that are used in the
on-line store.
20. The method of claim 19, wherein the query is created using
terms in the name of the product, the classification of the
product, the identification code of the product, and the
description of the product.
Description
CROSS-REFERENCE TO RELATED PATENT APPLICATIONS
[0001] This application is a continuation application under 35
U.S.C. .sctn. 365(c) of International Application No.
PCT/KR2006/001858, filed May 18, 2006 designating the United
States. International Application No. PCT/KR2006/001858 was
published in English as WO2006/0126800 A1 on Nov. 30, 2006. This
application further claims the benefit of the earlier filing date
under 35 U.S.C. .sctn. 365(b) of Korean Patent Application No.
10-2005-0044404 filed May 26, 2005. This application incorporates
herein by reference the International Application No.
PCT/KR2006/001858 including the International Publication No.
WO2006/0126800 A1 and the Korean Patent Application No.
10-2005-0044404 in their entirety.
BACKGROUND
[0002] 1. Field
[0003] The present disclosure relates to a price comparison engine
and more particularly to building and updating product databases
for a price comparison engine.
[0004] 2. Background
[0005] Generally, a price comparison engine maintains a product
database, which contains information of products offered by various
on-line stores. The information includes product names,
identification codes for products, product prices, product
descriptions, on-line stores offering products, and classifications
of products. Typically price comparison engines maintain their own
product classification system or adopt a publicly available product
classification system for facilitating product searches and
database management. For example, a cellular phone, a computer, and
an MP3 player are respectively classified like appliance/cellular
phone, computer/peripheral device, and appliance/audio/MP3. Also,
multimedia files, news, and other information may be classified and
stored along with their classification or category information.
Generally, the product category and classification system has a
hierarchical structure.
[0006] A shopping mall relay service or price comparison service
provides its users with price information for products and on-line
store information selling the products. When displaying the price
information, the service provides its users with a link of the
on-line stores. For providing the goods relay service, information
received from on-line stores is stored in a database in a goods
relay server.
[0007] The foregoing discussion in this section is to provide
general background information and does not constitute an admission
of prior art.
SUMMARY
[0008] One aspect of the invention provides a method of updating a
product database. The method comprises: providing a database
comprising a plurality of database entries, wherein each database
entry comprises a classification field and a plurality of
additional fields, wherein the classification field comprises a
classification of an item of the database entry; providing data
associated with a product item for adding to the database as a new
entry; generating a query from the data associated with the
product; computing a correlation value of the query with
information saved in at least one selected field of a database
entry, wherein computing is repeated for at least part of the
plurality of database entries; identifying database entries
including information in the at least one selected field, which has
a correlation value greater than a predetermined value; determining
the most popular classification among classifications saved in the
classification fields of the identified database entries; and
assigning the most popular classification to the classification
field of the new entry.
[0009] In the foregoing method, the data may be retrieved from an
on-line store for selling the product. The query may comprise a
word or phrase that the on-line store uses to identify, describe or
classify the product. The plurality of fields may further comprise
a product source field, wherein the method may further comprise
adding to the product source field a name of the on-line store. The
plurality of fields may further comprise a price field, wherein the
method may further comprise adding to the price field a price of
the product available at the on-line store. The method may further
comprise filtering the price field of a group of database entries
using a price range so as to select a smaller group of entries with
a price falling in the price range.
[0010] Still in the foregoing method, providing data may comprise
receiving the data from an on-line store, wherein the data may
comprise at least one selected from the group consisting of a name
of the product, a classification of the product, an identification
code of the product, and a description of the product that are used
in the on-line store. The query may be created using terms in the
name of the product, the classification of the product, the
identification code of the product, and the description of the
product. The plurality of fields may comprise a product
identification code field, wherein the method may further comprise
determining whether the database comprises a database entry having
an internal identifier in the product identification code field
that is identical to the identification code of the product that is
used in the on-line store. Generating, computing, identifying and
determining may be conducted only when it is determined that there
is no database entry that includes the identical identification
code in the product identification code field. The at least one
selected field may comprise the product identification code field.
The at least one selected field may comprise a field other than the
product identification code field. The internal identifier may be a
model number of the product.
[0011] Yet in the foregoing method, the query may comprise a string
of plurality of words associated with the product. Determining may
comprise: identifying different classifications used in the
identified database entries; counting the number of database
entries for each different classification; and comparing the
counted numbers to determine the most popular classification.
[0012] Another aspect of the invention provides a method of running
a price comparison engine, the method comprising: maintaining a
product database comprising the foregoing method; receiving, from a
remote computer, a search request for a type of products;
conducting a search in the database using information included in
the search request; and sending a search result to the remote
computer.
[0013] A further aspect of the invention provides a method of a
product database, which comprise: providing a product database
comprising a plurality of fields for each database entry, the
plurality of fields comprising a classification field; providing
data associated with a product for adding to the database as a new
entry; extracting at least one term from the data associated with
the product; conducting at least one search using the at least one
term in the database so as to locate a plurality of entries
including the at least one term in at least one field thereof; and
assigning, to the classification field of the new entry,
information or a value that is the most popular in the
classification fields of the plurality of entries located in the at
least one search.
[0014] In the foregoing method, the data may be retrieved from an
on-line store for selling the product. The data may comprise at
least one selected from the group consisting of a name of the
product, a classification of the product, an identification code of
the product, and a description of the product that are used in the
on-line store. The query may be created using terms in the name of
the product, the classification of the product, the identification
code of the product, and the description of the product.
[0015] An aspect of the present invention provides a category
recommendation method and a system which can extract a query
associated with a foreign identifier and recommend an internal
category having a high similarity with the query as a category
corresponding to the foreign category, so that a proper category
for the foreign category is effectively recommended.
[0016] An aspect of the present invention also provides a category
recommendation method and a system which can immediately and
accurately recommend a proper category for a foreign category from
internal categories stored in a database, so that information on
Internet websites is systemically collected.
[0017] An aspect of the present invention also provides a category
recommendation method and a system which can recommend a proper
category for a foreign identifier from an internal category stored
in a database, so that category matching is automated by increasing
a recommendation matching rate.
[0018] An aspect of the present invention also provides a category
recommendation method including: maintaining a category database
including an internal identifier and an internal category to which
the internal identifier belongs; externally receiving a foreign
identifier and a foreign category where the foreign identifier
belongs to from an outside; extracting a query associated with the
foreign identifier; and recommending at least one of the internal
categories stored in the category database by using the query as a
category corresponding to the foreign identifier.
[0019] In this case, the internal identifier and internal category
indicate an identifier and a category used in a predetermined goods
relay server, and the foreign identifier and foreign category
indicate an identifier and a category used in other goods relay
servers, excluding the predetermined goods relay server. In this
case, the identifier is for identifying goods, a multimedia file,
knowledge, news and the like, i.e. a goods name, a file name, a
keyword of the knowledge and a title of the news.
[0020] As an example, the internal identifier and the internal
category are an identifier and a category used in the predetermined
goods relay server, the foreign identifier and foreign category are
an identifier and a category used in a shopping mall server
associated with the goods relay server.
[0021] The extracted query may be used as a query for searching a
category database. In this case, both the foreign identifier and
the internal identifier may be goods identifiers, the recommending
at least one of the internal categories stored in the category
database by using the query as a category corresponding to the
foreign identifier may search the category database using the
extracted query and recommend internal identifiers whose similarity
to the query is greater than a predetermined level.
[0022] An aspect of the present invention provides a category
recommendation system including: a category database including an
internal identifier and an internal category to which the internal
identifier belongs; a receiving unit receiving a foreign identifier
and a foreign category to which the foreign identifier belongs; a
query extraction unit extracting a query associated with the
foreign identifier; and a category recommendation unit recommending
at least one of the internal category stored in the category
database by using the query as a category corresponding to the
foreign identifier.
BRIEF DESCRIPTION OF THE DRAWINGS
[0023] FIG. 1 is a diagram briefly illustrating a network
connection association of a goods relay server, a shopping mall
server, and goods register, according to an embodiment of the
present invention;
[0024] FIG. 2 is a flowchart illustrating operations of a category
recommendation method according to an embodiment of the present
invention;
[0025] FIG. 3 is diagram illustrating an embodiment of the category
database;
[0026] FIG. 4 is a flowchart illustrating an embodiment of the
category recommendation operation illustrated in FIG. 2;
[0027] FIG. 5 is a flowchart illustrating another embodiment of the
category recommendation operation illustrated in FIG. 2;
[0028] FIG. 6 is a diagram illustrating a display of counting a
number of internal identifiers for each internal category according
to a query, according to an embodiment of the present
invention;
[0029] FIG. 7 is a diagram illustrating a display of counting a
number of internal identifiers for each internal category according
to a query, according to another embodiment of the present
invention; and
[0030] FIG. 8 is a block diagram illustrating a configuration of a
category recommendation system according to an embodiment of the
present invention.
DETAILED DESCRIPTION OF EMBODIMENTS
[0031] Embodiments of the invention will be described in detail
with reference to the accompanying drawings.
[0032] In order to sell products, on-line store services provide
product information. Each on-line store classifies product
information into a category according to its own classification
system. For example, on-line store `A` classifies an accessory
product into the category of `fashion/jewelry/18K`, while on-line
store `B` classifies the same product into a category of
`accessory/gold/18K`.
[0033] Because these on-line stores' classifications can be
different from another, it would be difficult that a price
comparison engine's own product database adopts one of the various
classifications from the on-line stores. Thus, a price comparison
engine typically has its own classification system for use in its
product database. Various aspects of the invention involve building
and updating a product database for a price comparison engine. More
particularly, various features of the invention involve creating,
suggesting or proposing a classification of a product under the
classification system of the price comparison engine. Creating,
suggesting or proposing a classification utilizes the existing data
of the product database of the price comparison engine.
[0034] FIG. 1 is a diagram briefly illustrating a network
connection association of a goods relay server 100, a shopping mall
server 110, and goods register 120, according to an embodiment of
the present invention. For convenience of description, only the
shopping mall server 110 corresponding to the goods relay server
100 is illustrated in FIG. 1, however the goods relay server 100
may operate in interoperation with a plurality of the shopping mall
server 110.
[0035] A category recommendation method according to one embodiment
of the present invention may be performed by a predetermined
category system, and the category recommendation method may be
performed in the goods relay server 100 and the shopping mall
server 110, or in interoperation with the goods relay server 100
and the shopping mall server 110.
[0036] The category recommendation system maintains a category
database including an internal identifier, i.e. a goods identifier,
used in a predetermined goods relay server 100 and an internal
category, i.e. a goods category, in which the internal identifier
belongs, and receives a foreign identifier, i.e. a goods
identifier, and a foreign category, i.e. a goods category, of goods
information registered in the shopping mall server 110.
[0037] Also, the category recommendation system extracts a query
associated with the foreign identifier. In this case, the category
recommendation system may extract the query by exclusively using
the foreign identifier, i.e. a goods identifier, and may extract
the query by using both the foreign identifier and the foreign
category in which the foreign identifier belongs. The above
described extracting of the query may be performed by various
methods, and as long as the query associated with the foreign
identifier is extracted, any method is used for the extracting of
the query, is included in one embodiment of the present invention.
As an example, the query may be extracted by eliminating a space of
the foreign identifier, or may be extracted by using spacings of
the foreign identifier. Also, the query may be extracted by
eliminating the space of the foreign identifier and the foreign
category, by using the spacings of the foreign identifier and the
foreign category, or by analyzing a morpheme of the foreign
identifier and the foreign category. In this case, when the foreign
category has a hierarchical structure, the extracting of the query
extracts the query by using a partial field of the hierarchical
structure of the foreign category, i.e. a lowest hierarchical
category, where, for example, `18K` within `fashion/jewelry/18K`
may be used for the extracted query.
[0038] Next, the category recommendation system recommends at least
one of the internal categories stored in the category database by
using the query as a category corresponding to the foreign
identifier. In this case, the category recommendation system
automatically matches a most appropriate internal category to a
category corresponding to the foreign identifier.
[0039] Usually, when identifiers, i.e. a goods identifier, are
similar, categories in which the identifiers belong are accordingly
similar, so that a category for a foreign identifier, e.g. a new
goods identifier, may be immediately and accurately established by
searching the category database using the query, extracting
internal identifiers whose similarity to the query is greater than
a predetermined level, and recommending at least one of the
internal category as a category corresponding to the foreign
identifier, in an order of the internal category having a greater
number of extracted internal identifiers.
[0040] FIG. 2 is a flowchart illustrating operations of a category
recommendation method according to an embodiment of the present
invention.
[0041] The category recommendation method according to one
embodiment of the present invention may be performed by a
predetermined category recommendation system, and the category
recommendation system may be performed in a goods relay server and
a shopping mall server, or in interoperation with a goods relay
server and a shopping mall server.
[0042] Referring to FIG. 2, the category recommendation method
according to one embodiment of the present invention maintains a
category database in operation S210.
[0043] FIG. 3 is diagram illustrating an embodiment of the category
database.
[0044] As illustrated in FIG. 3, the category database may include
internal identifiers 301 used in a goods relay server and the like,
an internal category 302 in which each of the internal identifiers
301 belong, and a search field 303.
[0045] As an example, the internal identifier 301 may be a goods
identifier and an identifier (ID) used for identifying each goods
model in the goods relay server. The internal category 302 may be
an internal category used in the goods relay server. The internal
category 302 may be stored in a type of an ID indicating a
category. The search field 303 may be a field of a search object
when searching a recommendation category using a query. The search
field 303 may include information associated with each goods model,
e.g. a brand name, a model name, an item name, a keyword, and the
like, or information regarding a category. As an example, the
search field 303 may exclusively include information regarding the
internal identifier 301, and may include information regarding both
the internal identifier 301 and information regarding the internal
category 302.
[0046] In FIG. 3, the internal category is recorded in the category
database as a separate data field from the search field 303, and
which falls within the scope of the present invention. The internal
category 302 may be included in the search field 303 to be
recorded, not in the separate data field.
[0047] In operation S220, the category recommendation method
externally receives a foreign identifier and a foreign category to
which the external foreign identifier belongs from an outside.
[0048] As an example, the foreign identifier may be a goods
identifier. The foreign identifier and foreign category may be used
in the goods relay server and the shopping mall server, and each of
the shopping mall servers may use a different identifier for an
identical goods and a different category for the identical goods.
As above described, as long as the categories are different in each
of the shopping mall servers, the categories can not be registered
in the goods relay server, therefore appropriate categories for
various goods are required to be established and registered in the
goods relay server.
[0049] In operation S230, the category recommendation method
determines whether there is an internal identifier corresponding to
a received foreign identifier.
[0050] Namely, the category recommendation method determines
whether there is an internal identifier corresponding to a foreign
identifier received from the shopping mall server and the like,
from the internal category used in the goods relay server and the
like when both the foreign identifier and the internal identifier
are goods identifiers.
[0051] In operation S240, when there is the internal identifier
corresponding to the received foreign identifier, a category to
which the internal identifier belongs is selected as a category
corresponding to the received foreign category.
[0052] In operation S250, a query associated with the received
foreign identifier is extracted when there is no internal
identifier corresponding to the received foreign identifier, from
the internal category.
[0053] In this case, only the query may be extracted by exclusively
using the foreign identifier. When the query is extracted by
exclusively using the foreign identifier, a configuration of the
category recommendation system may be simplified and time spent for
extracting the query may be reduced. Also, the query may be
extracted by using a foreign identifier and a foreign category. The
extracting of the query by using the foreign identifier and the
foreign category may contribute to calculating a similarity among
information between the foreign identifier and the category
database, for example, in a case of a foreign category having a
three stage hierarchical structure, the query may be extracted by
using a foreign identifier and a highest, a middle and a lowest
hierarchy of a foreign category when extracting a query.
[0054] In this case, the extracting of the query may be performed
by various methods. As an example, the query may be extracted by
eliminating a space of the foreign identifier and/or foreign
category, or may be extracted by using spacings of the foreign
identifier and/or foreign category. Also, the query may be
extracted by analyzing a morpheme of the foreign identifier and/or
the foreign category. Also, the query may be extracted by
sequentially executing the above mentioned methods when satisfying
a predetermined search condition. As an example, when an internal
identifier is retrieved by using the extracted query, extracted by
eliminating a space from a foreign identifier and/or a foreign
category, and when a search result is not a desirable condition,
i.e. not as many as a desired number, the query is repeatedly
extracted by using a spacings from the foreign identifier and/or
the foreign category, and after the repeatedly extracting the
query, when a search result is not still a desirable condition,
i.e. not as many as a desired number, the query is repeatedly
extracted by analyzing a morpheme of the foreign identifier and the
foreign category.
[0055] In S260, a category recommendation method according to an
embodiment of the present invention recommends at least one of the
internal categories stored in the category database by using the
query as a category corresponding to the foreign identifier.
[0056] In this case, a most appropriate category may be exclusively
recommended, and the one recommended category may be automatically
established as a category corresponding to a received foreign
category.
[0057] Also, the recommended category may be a predetermined number
of categories according to the establishment. In this case, the
recommended category may be one category selected from a number of
categories recommended by a user registering the foreign
identifier, e.g. a goods identifier.
[0058] FIG. 4 is a flowchart illustrating an embodiment of the
category recommendation operation S260 illustrated in FIG. 2.
[0059] Referring to FIG. 4, in operation S410, the category
recommendation method searches the category database using the
query, and extracts internal identifiers whose similarity to the
query is greater than a predetermined level.
[0060] In this case, the similarity may be measured by various
methods. The scope of the present invention even includes when the
similarity is calculated by arbitrary methods, as long as a
similarity degree between information of the extracted query and
the category database is measured. As an example, the similarity
calculation may be identical to a calculated similarity included in
the paper `Optimizing the Weight of Added Terms in Query Expansion`
from 241 to 246 pages in a collection of learned papers of the
9.sup.th Korea Information Management Society Scholarship
conference.
[0061] In operation S410, an internal identifier whose similarity
is greater than 0.5 may be extracted.
[0062] In the extracting of the internal identifiers in the
operation S410 illustrated in FIG. 4, only the internal
identifiers, whose price corresponds to the internal identifiers
that are within a predetermined range from a price corresponding to
the foreign identifier, may be extracted. As an example, when a
price corresponding to a foreign identifier is 45,000 Won, a price
ranging from 35,000 to 55,000 Won is established as a range of a
price corresponding to internal identifiers to be extracted.
[0063] Similar goods within an identical or a similar category
normally have a similar range of a price, so that identifiers whose
similarity to the foreign identifier is high may be immediately and
easily extracted by limiting the range of the price.
[0064] In operation S420, the category recommendation operation
recommends at least one of the internal categories as a category
corresponding to the foreign identifier, in an order of the
internal category having a greater number of extracted internal
identifiers. FIG. 5 is a flowchart illustrating another embodiment
of the category recommendation operation S260 illustrated in FIG.
2.
[0065] A category database may include a search field generated by
using an internal identifier and an internal category.
[0066] Referring to FIG. 5, in operation S510, the category
recommendation operation searches a category database by using a
query, and extracts internal identifiers whose similarity to the
query is greater than a predetermined level.
[0067] As described above, the internal identifiers are extracted
based on the search field generated by using the internal
identifiers and internal category and the similarity among the
extracted queries, therefore a more appropriate category for the
foreign category may be recommended. In this case, the similarity
may be measured by various methods, e.g. in operation S510, an
internal identifiers whose similarity between the search field and
the query is greater than 0.5 may be extracted.
[0068] In the extracting of the internal identifiers in operation
S510 illustrated in FIG. 5, only internal identifiers whose a price
corresponding to internal identifiers are within a predetermined
range, from a price corresponding to a foreign identifier, may be
extracted.
[0069] In operation S520, the category recommendation operation
recommends at least one of the internal categories as a category
corresponding to the foreign identifier, in an order of the
internal category having a greater number of extracted internal
identifiers.
[0070] In the recommending of the internal category in operation
S260 illustrated in FIG. 5, the query may be extracted in another
method according to the foreign category and a server transmitting
the foreign category. Namely, when the foreign identifier is a
goods identifier and the server transmitting the foreign identifier
and foreign category is a shopping mall server, a query for each
shopping mall may be extracted by a different method. As an
example, for an A shopping mall server, a query is extracted by
using a goods identifier, for a B shopping mall server, a query is
extracted by using a lowest hierarchy of a foreign category and a
goods identifiers, and for a C shopping mall server, a query is
extracted by using an entire foreign category and goods identifier.
Therefore, the category may be effective by establishing an
appropriate recommendation method of extracting a query according
to a property of the server transmitting the foreign identifier and
the foreign category.
[0071] Furthermore, in the recommendation of the category in
operation S260 illustrated in FIG. 2, an object of a similarity to
the query may be different according to the foreign identifier and
the server transmitting the foreign category. As an example, for an
A shopping mall server, an internal identifier may be extracted
based on a similarity between the query and the internal
identifier, and for a B shopping mall server, an internal
identifier may be extracted based on a similarity among search
fields generated by using the query, the internal identifier, and
the internal category.
[0072] FIG. 6 is a diagram illustrating a display of counting a
number of internal identifiers for each internal category according
to a query, according to an embodiment of the present
invention.
[0073] Referring to FIG. 6, a foreign identifier 610 `loveletter
coconut`, and a foreign category 620
`food/mart@importedfood@snack(efoodmart)`, are received from an
outside shopping mall site, i.e. `dnshop`.
[0074] The category recommendation method according to one
embodiment of the present invention extracts queries 631, 632 and
633 associated with the foreign identifier 610 `loveletter coconut`
and the foreign category 620
`food/mart@importedfood@snack(efoodmart)`. In this case, the
queries 631, 632 and 633 associated with the foreign identifier 610
may be generated by exclusively using the foreign identifier 610 or
by using both the foreign identifier 610 and the foreign category
620. The query 633 is generated by exclusively using the foreign
identifier 610, and the queries 631 and 632 are generated by using
both the foreign identifier 610 and the foreign category 620.
Specifically, the query 631 is generated by using a lowest
hierarchy `snack(efoodmart)`from the foreign identifier 610 and the
foreign category 620, and the query 632 is generated by using all
hierarchies `food/mart@importedfood@snack(efoodmart)`from the
foreign identifier 610 and the foreign category 620.
[0075] Referring to a search result 641 that an internal identifier
for each internal category is extracted by using the query 631, a
number of an internal identifier whose similarity to the query 631,
including category ID `08040103`, within the internal category, is
greater than a predetermined level, i.e. over 0.5, is one.
[0076] Referring to a search result 642 that an internal identifier
for each internal category is extracted by using the query 632, a
number of an internal identifier whose similarity to the query 632,
including category ID `08040103`, within the internal category, is
greater than a predetermined level is one.
[0077] Referring to a search result 643 that an internal identifier
for each internal category is extracted by using the query 633, a
number of an internal identifier whose similarity to the query 633,
including category ID `08040103`, within the internal category, is
greater than a predetermined level is one.
[0078] As described above, identifier in one embodiment in FIG. 6,
using each of the queries 631, 632 and 633, the internal category
including a result of the extracting of the internal category is
one, therefore a category
`food/supermarket@processedfood@confectionaries@cookie/biscuit/p-
ie` is recommended as a category corresponding to the foreign
identifier 610.
[0079] In order to recommend an optimal internal category
corresponding to the foreign identifier 610, the queries 631, 632
and 633 may change, comparison objects for the similarity within a
category database to be searched by using the queries 631, 632 and
633 may be changed, or the similarity, i.e. a criterion of the
extracting of the internal identifier, may be changed. Furthermore,
as described above, according to a foreign identifier and a server
transmitting the foreign category, the comparison objects for the
similarity within the category database to be searched by using the
queries 631, 632, and 633, or the similarity, i.e. a criterion of
the extracting of the internal identifier, may be changed.
[0080] FIG. 7 is a diagram illustrating a display of counting a
number of internal identifiers for each internal category according
to a query, according to another embodiment of the present
invention.
[0081] Referring to FIG. 7, a foreign identifier 710
`[lulugold]royalcubic rolex couple ring cr01-53` and a foreign
category 720 `openmarket@fashion@jewelry@14K/18K` are received from
an outside shopping mall, i.e. interpark.
[0082] The category recommendation method according to one
embodiment of the present invention extracts queries 731, 732 and
733 associated with the foreign identifier 710 `[lulugold]
royalcubic rolex couple ring cr01-53`. In this case, the queries
731, 732 and 733 associated with the foreign identifier 710
`[lulugold] royalcubic rolex couple ring cr01-53` may be extracted
by exclusively using the foreign identifier 710 or by using both
the foreign identifier 710 and foreign category 720. In FIG. 7, the
query 733 is extracted by exclusively using the foreign identifier
710, and the queries 731 and 732 are extracted by using both the
foreign identifier 710 and foreign category 720. Specifically, the
query 731 is extracted by using a lowest hierarchy `14K/18K` from
the foreign category 720, and the query 732 is extracted by using
all hierarchies `openmarket@fashion@jewelry@14K/18K` from the
foreign category 720.
[0083] Referring to a search result 741 of an internal identifier
for each internal category that is extracted by using the query
731, a number of internal identifiers whose similarity to the query
731, including a category ID `14000003` within the internal
category, and is greater than a predetermined level, i.e. over 0.5,
is three.
[0084] A number of internal identifiers whose similarity to the
query 731, including a category ID `14000001` within the internal
category, is greater than a predetermined level, is three.
[0085] Also, a number of internal identifiers whose similarity to
the query 731, including a category ID `14000403` within the
internal category, is greater than a predetermined level, is
two.
[0086] Also, a number of internal identifiers whose similarity to
the query 731, including a category ID `14000000` within the
internal category, is greater than a predetermined level, is
two.
[0087] Also, a number of internal identifiers whose similarity to
the query 731, including a category ID `14000401` within the
internal category, is greater than a predetermined level, is
one.
[0088] Therefore, the category recommendation method according to
one embodiment of the present invention recommends a category
`accessory/watch@gold/siver@14K@couple ring` that includes a
greatest number of the internal identifiers whose similarity to the
query 731 is greater than a predetermined level, or a category
corresponding to `[lulugold] royal cubic rolex couple ring
CR01-53`. When the internal categories include an identical number
of internal identifier, selecting of a category may change
according to initial setting. As an example, an internal category
including a smaller ASCII code value may be recommended as a
category corresponding to the foreign identifier, from internal
categories including an identical number of internal
identifier.
[0089] Referring to the search result 742 that an internal
identifier is extracted for each internal category by using the
query 732, a number of internal identifiers whose similarity to the
query 732, including a category ID `14000003` within the internal
category, is greater than a predetermined level, i.e. over 0.5, is
three.
[0090] A number of internal identifiers whose similarity to the
query 732, including a category ID `14000403`, within the internal
category, is greater than a predetermined level, is two.
[0091] Also, a number of internal identifiers whose similarity to
the query 732, including a category ID `14070403` within the
internal category, is greater than a predetermined level, is
one.
[0092] Also, a number of internal identifiers whose similarity to
the query 732, including a category ID `14000001` within the
internal category, is greater than a predetermined level, is
one.
[0093] Also, a number of internal identifiers whose similarity to
the query 732, including a category ID `14000000` within the
internal category, is greater than a predetermined level, is
one.
[0094] Therefore, the category recommendation method according to
one embodiment of the present invention recommends the category
`accessory/watch@gold/siver@14K@couple ring` that includes a
greatest number of the internal identifiers whose similarity to the
query 732 is greater than a predetermined level, corresponding to
the foreign identifier 710 `[lulugold] royal cubic rolex couple
ring CR01-53`.
[0095] Referring to a search result 743 that an internal identifier
for each internal category is extracted by using the query 733, a
number of internal identifiers whose similarity to the query 733,
including a category ID `14000003` within the internal category, is
greater than a predetermined level, i.e. over 0.5, is three.
[0096] A number of internal identifiers whose similarity to the
query 733, including a category ID `14000403` within the internal
category, is greater than a predetermined level, is two.
[0097] Also, a number of internal identifiers whose similarity to
the query 733, including a category ID `14000401` within the
internal category, is greater than a predetermined level, is
one.
[0098] Also, a number of internal identifiers whose similarity to
the query 733, including a category ID `14000000` within the
internal category, is greater than a predetermined level, is
one.
[0099] Therefore, the category recommendation method according to
one embodiment of the present invention recommends the category
`accessory/watch@gold/siver@14K@couple ring` including a greatest
number of the internal identifiers whose similarity to the query
733 is greater than a predetermined level, corresponding to the
foreign identifier 710 `[lulugold] royal cubic rolex couple ring
CR01-53`.
[0100] In certain embodiments in FIG. 7, a query extraction method
may be changed in order to recommend a most appropriate internal
category.
[0101] As described above, when one internal category is
recommended, the one recommended internal category may be
automatically established as an internal category corresponding to
a received foreign category.
[0102] From the above, the case in which one internal category
including a greatest number of the internal identifiers whose
similarity to the query is greater than a predetermined level is
exclusively recommended, is described, however over two internal
categories may be recommended in an order of the internal category
having a greater number of internal identifiers. In this case, a
user registered a foreign identifier may select one from over the
recommended two categories. As an example, the category
recommendation method may recommend two internal categories
`accessory/watch@gold/siver@14K@couple ring` and
`accessory/watch@gold/siver@silver@couple ring` as a category
corresponding to the foreign category 710 `[lulugold] royalcubic
rolex couple ring cr01-53`.
[0103] Besides the changes of the queries 731, 732 and 733, in
order to recommend an optimal internal category corresponding to
the foreign identifier 710, comparison objects for the similarity
within a category database to be searched by using the queries 731,
732 and 733 may be changed, or the similarity, i.e. a criterion of
the extracting of the internal identifier, may be changed.
Furthermore, as described above, according to a foreign identifier
and a server transmitting the foreign category, the comparison
objects for the similarity within the category database to be
searched by using the queries 731, 732 and 733 or the similarity,
i.e. a criterion of the extracting of the internal identifier, may
be changed. As an example, in a case of a foreign identifier and a
foreign category received from the A shopping mall, a query is
extracted by using the foreign identifier, the foreign category,
and the query 731, in a case of a foreign identifier and a foreign
category received from the B shopping mall, a query is extracted by
using the foreign identifier and the query 733.
[0104] The category recommendation method according to one
embodiment of the present invention may be recorded in
computer-readable media including program instructions to implement
various operations embodied by a computer. The media may also
include, alone or in combination with the program instructions,
data files, data structures, and the like. Examples of
computer-readable media include magnetic media such as hard disks,
floppy disks, and magnetic tape; optical media such as CD ROM disks
and DVD; magneto-optical media such as optical disks; and hardware
devices that are specially configured to store and perform program
instructions, such as read-only memory (ROM), random access memory
(RAM), flash memory, and the like. The media may also be a
transmission medium such as optical or metallic lines, wave guides,
etc. including a carrier wave transmitting signals specifying the
program instructions, data structures, etc. Examples of program
instructions include both machine code, such as produced by a
compiler, and files containing higher level code that may be
executed by the computer using an interpreter. The described
hardware devices may be configured to act as one or more software
modules in order to perform the operations of one embodiment of the
present invention.
[0105] FIG. 8 is a block diagram illustrating a configuration of a
category recommendation system according to an embodiment of the
present invention.
[0106] Referring to FIG. 8, a category recommendation system 800
according to an embodiment of the present invention includes a
category database 801, a receiving unit 802, a query extraction
unit 803 and a category recommendation unit 804.
[0107] The category database 801 includes an internal identifier
and an internal category where the internal identifier belongs to.
In this case, the internal identifier and the internal category
indicate an identifier and a category used in a predetermined goods
relay server, which is distinguished from a foreign identifier and
a foreign category indicate identifiers and categories used in
other servers, excluding the server. In this case, the identifier
is for identifying goods, a multimedia file, knowledge, news, and
the like, i.e. a goods name, a file name, a keyword of the
knowledge, and a title of the news.
[0108] The category database 801 further includes a search field,
as described in FIG. 3. The internal category may be stored in a
type of an identifier indicating a category. The search field
indicates a field of a search object when searching a
recommendation category by using a query. The search field may
include information associated with a goods model, e.g. a brand
name, a model name, an item name, a keyword, and the like, or
information regarding a category. As an example, the search field
may exclusively include information regarding the internal
identifier, and may include both the information regarding the
internal identifier and information regarding the internal
category.
[0109] The receiving unit 802 receives a foreign identifier and a
foreign category to which the foreign identifier belongs from an
outside.
[0110] As an example, the query extraction unit 803 extracts a
query associated with the foreign identifier. In this case, when
the query is extracted by exclusively using a foreign identifier,
the configuration of a category recommendation system may be
simplified and time spent for extracting the query may be reduced.
As an example, in a case of a foreign category having a three stage
hierarchical structure, the query may be extracted by using a
lowest hierarchy of the foreign identifier and the foreign category
when extracting a query. Also, in the case of the foreign category
having the three stage hierarchical structure, the query may be
extracted by using a foreign identifier and a highest, a middle and
a lowest hierarchy of the foreign category when extracting a
query.
[0111] In this case, the extracting of the query may be performed
by various methods. As an example, the query may be extracted by
eliminating a space of the foreign identifier and/or foreign
category, or may be extracted by using spacings of the foreign
identifier and/or foreign category. Also, the query may be
extracted by analyzing a morpheme of the foreign identifier and/or
the foreign category. Also, the query may be extracted by
sequentially executing the above mentioned methods when satisfying
a predetermined search condition. As an example, when an internal
identifier is retrieved by using the extracted query through
eliminating a space from a foreign identifier and/or a foreign
category to extract a query, and when a search result is not in a
desirable condition, i.e. not as many as a desired number, the
query is repeatedly extracted by using spacings from the foreign
identifier and/or the foreign category, and after the repeatedly
extracting the query, when a search result is not still a desirable
condition, i.e. not as many as a desired number, the query is
repeatedly extracted by analyzing a morpheme of the foreign
identifier and the foreign category.
[0112] The category recommendation unit 804 recommends at least one
of the internal categories stored in the category database 801 by
using the query as a category corresponding to the foreign
identifier.
[0113] In this case, a most appropriate category may be exclusively
recommended and the one recommended category may be automatically
established as a category corresponding to a received foreign
identifier.
[0114] Also, the recommended category may be a predetermined number
of the category according to the establishment. In this case, the
recommended category may be one selected category of categories
recommended by a user registering the foreign identifier, e.g. a
goods identifier.
[0115] The category recommendation unit 804 may perform according
to a flow of operations described through FIGS. 4 and 5.
[0116] The category recommendation unit 804 searches the category
database 801 by using the query, and extracts internal identifiers
whose similarity to the query is greater than a predetermined
level; and recommends at least one of the internal categories as a
category corresponding to the foreign identifier, in an order of
the internal category having a greater number of extracted internal
identifiers.
[0117] Although embodiments of the present invention have been
shown and described, the present invention is not limited to the
described embodiments. Instead, it would be appreciated by those
skilled in the art that changes may be made to these embodiments
without departing from the principles and spirit of the invention,
the scope of which is defined by the claims and their
equivalents.
[0118] According to embodiments of the present invention, there is
provided a category recommendation method and a system which can
extract a query associated with a foreign identifier and recommend
an internal category having a high similarity with the query as a
category corresponding to the foreign category, so that a proper
category for the foreign category is effectively recommended.
[0119] According to embodiments of the present invention, there is
provided a category recommendation method and a system which can
immediately and accurately recommend a proper category for a
foreign category from internal categories stored in a database, so
that information on Internet websites is systemically is
collected.
[0120] According to embodiments of the present invention, there is
provided a category recommendation method and a system which can
recommend a proper category for a foreign identifier from an
internal category stored in a database, so that category matching
is automated by increasing a recommendation matching rate.
* * * * *